John Healey’s first Budget on 28 October looks particularly nightmarish. It isn’t just that the Iran war has pushed up bond yields, potentially costing up to half the £21.7 billion headroom left behind by Rachel Reeves. But new governments mean new priorities. His own is defence, where he publicly called on Sir Keir Starmer to raise spending to 3% of GDP by 2030, even resigning over the issue. That would cost £10-15 billion, the equivalent of nearly 2p on income tax. But at the same time his new boss, Andy Burnham, has big ambitions somewhere else, namely social care. And that could cost the same again if, for example, reforms included introducing Scottish-style free personal care and a pay rise to put social care workers on the same level as their NHS equivalents.
The Chancellor will no doubt ask Treasury officials to hunt around for politically-bearable tax rises or spending cuts. The trouble is those same civil servants have been doing just that for Rachel Reeves for two years (and before that for me). When you rule out the three big taxes, there is little else to tax except business. That cripples growth – and Rachel Reeves paid a heavy price as a result. He may look at new taxes on banks. But given how mobile international capital is, it would be highly risky and probably end up losing money.
Then there are savings. The most obvious choice is welfare reform. Restoring working-age welfare spend to 2019 levels in real terms would save £56 billion a year within five years. But that has been kicked to the long grass and most people think that it is unlikely Andy Burnham will want to challenge his own MPs on the issue. Didn’t he tell them he was going to use the whip more lightly?
But there is one other area with real potential for savings. I hope it won’t send you to sleep if I utter the words ‘public sector productivity’ – because improving it offers the chance to fund a big rise in defence spending without damaging tax rises. That’s because if we made the public sector more efficient, we could increase spending on it more slowly, allowing the savings to be diverted to our armed forces.
The public sector represents a fifth of the UK economy, so its productivity matters to our overall performance. Yet according to the Institute for Fiscal Studies, it declined by an average of 0.3 per cent a year between 1997 and 2019. It has recovered since, but still remains below pre-pandemic levels. That directly links to Britain’s debt problem: when services require more funding without becoming more efficient, persistent deficits become harder to avoid.
But the opposite is also true. Remember those scary figures from the Office for Budget Responsibility saying our national debt is projected to be around 270% of GDP by the 2070s? When I saw those figures as Chancellor, I asked the Treasury how much public sector productivity would need to improve to avoid such a cataclysmic outcome. I published the analysis they did and it was fascinating. It showed that raising public sector productivity by only 0.7 per cent a year would stabilise national debt as a proportion of GDP at current levels. Doing better than 0.7% could actually reduce it.
It may sound ambitious but I am certain it can be done. Long-term productivity growth in the private sector is 2%. Of course it’s harder in the public sector where things are more politicised and unions are strong. But one relatively straightforward measure – bringing Civil Service numbers down to pre-pandemic levels – would save as much as £5 billion a year.
Promises by politicians to deliver savings are rarely believed. I am painfully aware of this because my own spending plans, which included ambitious savings targets, were dismissed as “fiscal fiction”. This time, however, there are two big reasons why it could be different.
The first is AI. Microsoft estimates AI could add 2 per cent a year to productivity and growth. They have skin in the game, so you may prefer a more cautious estimate from Nobel Prize-winning economist Philippe Aghion, who puts the figure at around 1 per cent. Either way, the Treasury’s 0.7 per cent target (which would include lots of other efficiency savings) starts to look more achievable.
The second reason is that the UK’s biggest public service, the NHS, is already delivering 2% productivity improvements. In my 2024 budget I gave the NHS £3.4 billion for upfront technology investment in return for a productivity commitment of 2 per cent a year. Labour kept those plans in place and the latest NHS England figures suggest that the NHS has more than delivering on its side of the bargain. In fact the NHS is starting to look like one of the most AI-enabled health services anywhere thanks to rapid adoption of new technologies such as ambient scripting which save massive amounts of admin. In other parts of the public sector there has been even more radical change, for example at the Passport Office which is now recognised as one of the most efficient in the world. Both the US government and Taiwan have sent delegations to study what it has done.
Rolling out such a public sector productivity plan would need to avoid some familiar Whitehall bear traps. Any AI transformation will need capital spending up front, something the Treasury is famously reluctant to release. We would also need to overhaul the way productivity is measured. Shorter NHS waiting times and lower crime rates are important public benefits but don’t release financial savings. But for any impact on public finances, improvements need to include cash savings, something we don’t currently measure very well.
Adopting a 2 per cent annual productivity target across the public sector for the next decade would be a major endeavour. It would need the full support of the Prime Minister, the Chancellor and the Cabinet Secretary. But the prize is huge. Taking the plunge would not just release a lot of money for defence or social care. It would make our creaking public sector far more responsive. It would turn the UK into a global leader for AI adoption in public services. And most importantly it would tackle our longest-standing barrier to growth, low productivity. The one fifth of our output represented by the public sector is also the part over which ministers have most direct control. Why not start there?
A shortened version of this article appeared in the Financial Times on Monday 24th August
Burnham: “the right way to fund defence.”
Interesting timing: at PMQs today, the Prime Minister argued that increased defence spending could partly be funded by building more council homes to reduce the housing-benefit bill, alongside education and mental-health measures intended to lower welfare dependency.
This is not quite public-sector productivity in the narrower sense described here, but it rests on a similar principle: invest upfront to reduce recurring public expenditure later. The unanswered question is the one your article rightly identifies—when do these policies generate actual cash savings, of what scale, and will those savings arrive soon enough to fund the defence commitments? Without quantified timing and cash-release targets, it remains an aspiration rather than a funded plan.
My view is that there is a clear and present danger for Defence - and we dont have 5 years to plan....
Another disingenuous political criminal and proven liar